Article
Computer Science, Artificial Intelligence
Kangjia Qiao, Kunjie Yu, Boyang Qu, Jing Liang, Hui Song, Caitong Yue
Summary: This article presents an evolutionary multitasking-based constrained multiobjective optimization framework for solving CMOPs. It transforms the optimization problem into two related tasks and utilizes a tentative method to discover and transfer useful knowledge. The approach achieves better performance compared to other state-of-the-art algorithms.
IEEE TRANSACTIONS ON EVOLUTIONARY COMPUTATION
(2022)
Article
Computer Science, Artificial Intelligence
Kangjia Qiao, Kunjie Yu, Boyang Qu, Jing Liang, Hui Song, Caitong Yue, Hongyu Lin, Kay Chen Tan
Summary: This article proposes a new multitasking-constrained multiobjective optimization (MTCMO) framework to solve complex CMOPs by creating a dynamic auxiliary task. The auxiliary task's constraint boundary dynamically reduces to maintain high relatedness with the main task and continuously provide supplementary evolutionary directions. An improved method is designed for the auxiliary task to utilize diverse high-quality infeasible solutions, and a new test function is introduced to analyze the characteristics of MTCMO. Compared with 11 state-of-the-art peer methods, the superior or competitive performance of MTCMO is demonstrated on benchmark functions and real-world applications.
IEEE TRANSACTIONS ON EVOLUTIONARY COMPUTATION
(2023)
Article
Computer Science, Artificial Intelligence
Zedong Tang, Maoguo Gong, Yue Wu, Wenfeng Liu, Yu Xie
Summary: The article presents a novel and computationally efficient intertask information transfer strategy by aligning subspaces. By introducing a learnable alignment matrix, it extracts complementary information among different tasks to enhance the performance of solving complicated problems. This method shows superior performance compared to existing evolutionary multitask optimization algorithms in comprehensive experiments.
IEEE TRANSACTIONS ON EVOLUTIONARY COMPUTATION
(2021)
Article
Computer Science, Information Systems
Fuhao Gao, Weifeng Gao, Lingling Huang, Jin Xie, Maoguo Gong
Summary: This paper proposes an evolutionary multitasking optimization algorithm that transfers effective knowledge through semi-supervised learning. By utilizing labeled and unlabeled samples generated in the optimization process, the algorithm identifies individuals with valuable knowledge and transfers the knowledge between tasks, leading to significantly improved performance.
INFORMATION SCIENCES
(2022)
Article
Automation & Control Systems
Zhengping Liang, Hao Dong, Cheng Liu, Weiqi Liang, Zexuan Zhu
Summary: Evolutionary multitasking (EMT) operates in the search space of multiple optimization tasks simultaneously, enhancing task-solving abilities through knowledge sharing. A novel multiobjective EMT algorithm called MOMFEA-SADE, based on subspace alignment and self-adaptive differential evolution, demonstrates superior performance in experimental results and won a competition within IEEE 2019 Congress on Evolutionary Computation.
IEEE TRANSACTIONS ON CYBERNETICS
(2022)
Article
Automation & Control Systems
Kangjia Qiao, Jing Liang, Zhongyao Liu, Kunjie Yu, Caitong Yue, Boyang Qu
Summary: This paper proposes a method based on multi-task evolutionary algorithms to solve constrained multi-objective optimization problems (CMOPs). By creating a main task, a global auxiliary task, and a local auxiliary task, the objective functions are optimized and constraint conditions are satisfied, improving the effectiveness of the solution.
IEEE-CAA JOURNAL OF AUTOMATICA SINICA
(2023)
Article
Automation & Control Systems
Hao Sun, Pengfei Chen, Ziyu Hu, Lixin Wei
Summary: The superior performance of evolutionary multitasking algorithms is attributed to the potential synergy between tasks. Current algorithms only transfer individuals from the source to the target task in a unidirectional process. This approach fails to consider the target task's search preference, leading to underutilization of task synergy. We propose a bidirectional knowledge transfer method that takes into account the target task's search preference when finding transferred individuals. Experimental results show that the proposed algorithm outperforms other comparison algorithms in over 30 benchmarks and exhibits considerable convergence efficiency.
Article
Computer Science, Artificial Intelligence
Hao Xu, A. K. Qin, Siyu Xia
Summary: Evolutionary Multitask Optimization (EMTO) uses evolutionary algorithms (EAs) to solve multiple optimization tasks simultaneously, utilizing knowledge transfer to improve performance. The proposed adaptive EMTO (AEMTO) framework adjusts knowledge transfer in a synergistic way, effectively addressing negative knowledge transfer and enhancing overall performance.
IEEE TRANSACTIONS ON EVOLUTIONARY COMPUTATION
(2022)
Article
Automation & Control Systems
Zefeng Chen, Yuren Zhou, Xiaoyu He, Jun Zhang
Summary: In this article, an evolutionary multitasking algorithm with learning task relationships is proposed for multiobjective multifactorial optimization (MO-MFO). The algorithm models the decision spaces of different tasks as a joint manifold and utilizes a joint mapping matrix to transfer information across different decision spaces. Experimental results demonstrate its superior performance compared to other state-of-the-art solvers in tackling complex MO-MFO problems involving heterogeneous decision spaces.
IEEE TRANSACTIONS ON CYBERNETICS
(2022)
Article
Computer Science, Information Systems
Qianlong Dang, Weifeng Gao, Maoguo Gong, Shuai Yang
Summary: This study presents a multi-objective evolutionary multitasking algorithm based on the positive knowledge transfer mechanism. It introduces a cheap surrogate model to evaluate the quality of solutions, designs a diversity maintenance method to preserve solution diversity, proposes a selection strategy for transferred solutions to improve the efficiency of positive knowledge transfer, and demonstrates the effectiveness and competitiveness of the algorithm through experiments.
INFORMATION SCIENCES
(2022)
Article
Automation & Control Systems
Jiabin Lin, Hai-Lin Liu, Kay Chen Tan, Fangqing Gu
Summary: Multiobjective multitasking optimization (MTO) is a novel research topic in the field of evolutionary computation, aiming to solve multiple related multiobjective optimization problems simultaneously using evolutionary algorithms. The key lies in the knowledge transfer based on sharing solutions across tasks. This study proposes a new algorithm to address MTO problems and validates its effectiveness through numerical studies on benchmark problems.
IEEE TRANSACTIONS ON CYBERNETICS
(2021)
Article
Automation & Control Systems
Huangke Chen, Ran Cheng, Witold Pedrycz, Yaochu Jin
Summary: This paper proposes a method to solve multiobjective optimization problems through multi-stage evolutionary search, highlighting convergence and diversity in different search stages. The algorithm balances and addresses the issues in multiobjective optimization through two stages.
IEEE TRANSACTIONS ON SYSTEMS MAN CYBERNETICS-SYSTEMS
(2021)
Article
Computer Science, Artificial Intelligence
Kangjia Qiao, Jing Liang, Kunjie Yu, Minghui Wang, Boyang Qu, Caitong Yue, Yinan Guo
Summary: This paper proposes a double-balanced evolutionary multi-task optimization (DBEMTO) algorithm to better solve constrained multi-objective optimization problems (CMOPs). DBEMTO evolves two populations to solve the main task (CMOP) and the auxiliary task (MOP extracted from the CMOP) respectively and uses three evolutionary strategies for offspring generation. DBEMTO has performed more competitively compared to other state-of-the-art CMOEAs according to the final results.
IEEE TRANSACTIONS ON EMERGING TOPICS IN COMPUTATIONAL INTELLIGENCE
(2023)
Article
Computer Science, Artificial Intelligence
Yinglan Feng, Liang Feng, Sam Kwong, Kay Chen Tan
Summary: The proposed multivariation multifactorial evolutionary algorithm aims to solve LSMOPs by conducting an evolutionary search on both the original space of the LSMOP and multiple simplified spaces constructed in a multivariation manner concurrently. This approach seamlessly transfers useful traits from simplified problem spaces to the original problem space, ensuring preservation of the original global optimal solution. Experimental results demonstrate the efficiency and effectiveness of the proposed method for large-scale multiobjective optimization compared to existing state-of-the-art methods.
IEEE TRANSACTIONS ON EVOLUTIONARY COMPUTATION
(2022)
Article
Computer Science, Artificial Intelligence
Zhengping Liang, Yingmiao Zhu, Xiyu Wang, Zhi Li, Zexuan Zhu
Summary: Evolutionary multitasking (EMT) is a promising topic in evolutionary computation where multiple related optimization tasks can be solved simultaneously through knowledge sharing. This article proposes EMT-GS, an EMT algorithm for multiobjective optimization, based on generative strategies using generative adversarial networks (GANs) and inertial differential evolution (IDE). The performance of EMT-GS is evaluated on three multitasking multiobjective benchmark problems, showing its competitiveness compared to other state-of-the-art multiobjective EMT algorithms.
IEEE TRANSACTIONS ON EVOLUTIONARY COMPUTATION
(2023)
Article
Computer Science, Artificial Intelligence
Shuling Yang, Kangshun Li, Zhengping Liang, Wei Li, Yu Xue
Article
Computer Science, Artificial Intelligence
Xiaoliang Ma, Xiaodong Li, Qingfu Zhang, Ke Tang, Zhengping Liang, Weixin Xie, Zexuan Zhu
IEEE TRANSACTIONS ON EVOLUTIONARY COMPUTATION
(2019)
Article
Computer Science, Information Systems
Zhengping Liang, Weijun Hou, Xiang Huang, Zexuan Zhu
INFORMATION SCIENCES
(2019)
Article
Computer Science, Information Systems
Zhengping Liang, Shunxiang Zheng, Zexuan Zhu, Shengxiang Yang
INFORMATION SCIENCES
(2019)
Article
Computer Science, Artificial Intelligence
Zhengping Liang, Jian Zhang, Liang Feng, Zexuan Zhu
EXPERT SYSTEMS WITH APPLICATIONS
(2019)
Article
Computer Science, Artificial Intelligence
Zhengping Liang, Ya Zou, Shunxiang Zheng, Shengxiang Yang, Zexuan Zhu
Summary: A novel feedback-based prediction strategy (FPS) with two feedback mechanisms is proposed to improve prediction accuracy and enhance the effectiveness of re-initialization. Experimental results demonstrate the effectiveness and efficacy of the proposed method in solving dynamic multi-objective optimization problems.
EXPERT SYSTEMS WITH APPLICATIONS
(2021)
Article
Automation & Control Systems
Zhengping Liang, Kaifeng Hu, Xiaoliang Ma, Zexuan Zhu
Summary: Balancing population diversity and convergence is crucial for evolutionary algorithms to solve many-objective optimization problems. The proposed two-round environmental selection strategy shows good performance in achieving this balance, as demonstrated through experiments.
IEEE TRANSACTIONS ON CYBERNETICS
(2021)
Article
Computer Science, Artificial Intelligence
Zhengping Liang, Jiyu Zeng, Ling Liu, Zexuan Zhu
Summary: This study proposed a new mutation strategy VCEM for many-objective evolutionary algorithms, which improves search efficiency by categorizing decision variables and guiding mutation based on different types of elite individuals. The algorithm was found to be more competitive than state-of-the-art MaOEAs on benchmark problems, demonstrating its effectiveness and generality.
SWARM AND EVOLUTIONARY COMPUTATION
(2021)
Article
Automation & Control Systems
Zhengping Liang, Tingting Luo, Kaifeng Hu, Xiaoliang Ma, Zexuan Zhu
Summary: This article introduces a new indicator-based many-objective evolutionary algorithm, MaOEA-IBP, with boundary protection to address the challenges faced by traditional multiobjective evolutionary algorithms when dealing with MaOPs. Experimental results demonstrate that MaOEA-IBP achieves competitive performance compared to other algorithms across various benchmark MaOPs.
IEEE TRANSACTIONS ON CYBERNETICS
(2021)
Article
Computer Science, Artificial Intelligence
Zhengping Liang, Xiuju Xu, Ling Liu, Yaofeng Tu, Zexuan Zhu
Summary: This article proposes an evolutionary many-task optimization algorithm, EMaTO-MKT, based on a multisource knowledge transfer mechanism. The algorithm adaptively determines the probability of using knowledge transfer and balances the self-evolution and knowledge transfer among tasks. It selects multiple highly similar tasks as learning sources and applies a knowledge transfer strategy based on local distribution estimation. Experimental results show the competitiveness of EMaTO-MKT in solving many-task optimization problems.
IEEE TRANSACTIONS ON EVOLUTIONARY COMPUTATION
(2022)
Article
Automation & Control Systems
Zhengping Liang, Tiancheng Wu, Xiaoliang Ma, Zexuan Zhu, Shengxiang Yang
Summary: In recent years, dynamic multiobjective optimization problems (DMOPs) have gained increasing attention. This article proposes a dynamic multiobjective evolutionary algorithm (DMOEA-DVC) based on decision variable classification, aiming to balance population diversity and convergence. Experimental results comparing DMOEA-DVC with six other algorithms on 33 benchmark DMOPs demonstrate its superior overall performance.
IEEE TRANSACTIONS ON CYBERNETICS
(2022)
Article
Automation & Control Systems
Zhengping Liang, Hao Dong, Cheng Liu, Weiqi Liang, Zexuan Zhu
Summary: Evolutionary multitasking (EMT) operates in the search space of multiple optimization tasks simultaneously, enhancing task-solving abilities through knowledge sharing. A novel multiobjective EMT algorithm called MOMFEA-SADE, based on subspace alignment and self-adaptive differential evolution, demonstrates superior performance in experimental results and won a competition within IEEE 2019 Congress on Evolutionary Computation.
IEEE TRANSACTIONS ON CYBERNETICS
(2022)
Article
Automation & Control Systems
Zhaobo Liu, Guo Li, Haili Zhang, Zhengping Liang, Zexuan Zhu
Summary: In this article, we propose a new multifactorial evolutionary algorithm based on diffusion gradient descent (MFEA-DGD) for multitasking optimization. The MFEA-DGD implements knowledge transfer among optimization tasks and ensures convergence through complementary crossover and mutation operators. Experimental results show that MFEA-DGD outperforms state-of-the-art EMT algorithms in terms of convergence speed and competitive results, and the convexity of different tasks provides interpretability of the experimental results.
IEEE TRANSACTIONS ON CYBERNETICS
(2023)
Proceedings Paper
Telecommunications
Guo Li, Ling Liu, Zhengping Liang, Xiaoliang Ma, Zexuan Zhu
Summary: Mobile edge computing (MEC) complements cloud computing by overcoming long physical transmission distances and accelerating edge computing servers. Implementing MEC in 5G networks ensures ultralow latency, but optimizing user service migration poses an NP-hard problem. By proposing the MA-CDLS algorithm, we aim to continually optimize service migration in 5G MEC scenarios, achieving lower user-perceived latency and energy consumption compared to traditional methods like profile tracking and game theory, particularly during congestion.
2021 IEEE 32ND ANNUAL INTERNATIONAL SYMPOSIUM ON PERSONAL, INDOOR AND MOBILE RADIO COMMUNICATIONS (PIMRC)
(2021)